Hello guys, well this time I come again with this game, Swordsman, for some reason the file is to big and when I try unpack them with old unpacker or script it say the file is to big and can't recognize pck, so my question is if possible help to unpack it... I know is possible got files with offzip, but ofc the files get out are wrong, why? well simple in theory the files was exported with random names and wrong extentions, a example .mox files they are not, because it use same format as perfect world, ski for models, stck for animatiosn, dds for textures and bon for bones, so well is possible check the format and try make a bms script for unpack them maybe? really this would be great because nobody do a release of this tool and with old tools you can't got files, many thanks to all community and have a nice day.
That's very strange because the models.pck splitted file you uploaded seems to work here and I'm able to extract it correctly. Are you sure that the sample you provided and this one are the same?
You can also start quickbms with the -V option to have a debug log of what field read from the file is wrong.
aluigi wrote:That's very strange because the models.pck splitted file you uploaded seems to work here and I'm able to extract it correctly. Are you sure that the sample you provided and this one are the same?
You can also start quickbms with the -V option to have a debug log of what field read from the file is wrong.
It looks like the filecutter script doesn't support 64bit fields but its content is correct. I have just updated the script so you can use it to split the archive again: http://aluigi.org/papers/bms/others/filecutter.bms (now it's 0.1.1)
The 200mb cut is ok.
For me it's impossible to download the original client because it's too slow.
ok here we go, I cut again in parts of 100MB so here cutted parts from new script anyway I test download part just in case and take a look, here I leave a link.
The reason of why the script worked on the 32bit-capped splits is funny, basically the script was reading just the 32bit part of the offset located in the archive and so that's why it didn't work. But it worked with the rebuilt 32bit split just for this reason
